The latest release of vesoft-inc/nebula is v3.8.0. It is an Apache-2.0 licensed open-source distributed graph database implemented in C++, aimed at storing and querying/analyzing large-scale data with many relationships. The repository description emphasizes horizontal scalability and high availability. However, the given records contain no changelog for v3.8.0, so what exactly changed in this version cannot be confirmed.
The core of this repository is its design that separates storage from compute. The project was created in August 2018 and has 12,439 stars. The README states that NebulaGraph processes large-scale data with millisecond-level latency while scaling quickly, and supports fast graph analysis. It provides symmetric distribution (a layout where every node in the cluster plays the same role), separation of storage and compute, horizontal scaling, strong data consistency based on the RAFT protocol, and an OpenCypher-compatible query language. Role-based access control and various graph analysis algorithms are included as well.
That said, the millisecond-level latency and scaling speed are only claims made by the README; the given information contains no performance figures backing them up and no comparison results against other graph databases. The research record's momentum score is a low 0.8 out of 10, and no novelty, technology, or quality scores are recorded. The repository name is vesoft-inc/nebula, while the README uses NebulaGraph as the product name, so it is worth knowing this naming difference before reading the documentation. For teams that work with data full of relationships, it is worth checking the query structure directly first.